problem
We keep getting NewRelic alerts on planningalerts exceeding 75% memory usage
Describe the solution you'd like
The system is just over the limit - some simple efficiencies should help it.
Describe alternatives you've considered
Increasing the memory usage trigger
Additional context
It doesn’t correlate to network or cpu usage spikes:
The planning alerts PR has significantly dropped memory usage, and the infrastructure PR gives it extra headroom (512 MB swap). I applied the changes manually to validate the syntax and approach. The first drop was restarting sidekiq and the second was restarting the 3 puma processes - they are now all monitored by systemd for MemoryMax
problem
We keep getting NewRelic alerts on planningalerts exceeding 75% memory usage
Describe the solution you'd like
The system is just over the limit - some simple efficiencies should help it.
Describe alternatives you've considered
Increasing the memory usage trigger
Additional context
It doesn’t correlate to network or cpu usage spikes:
The planning alerts PR has significantly dropped memory usage, and the infrastructure PR gives it extra headroom (512 MB swap). I applied the changes manually to validate the syntax and approach. The first drop was restarting sidekiq and the second was restarting the 3 puma processes - they are now all monitored by systemd for MemoryMax